March 26, 1998
fuer Landesentwicklung und Umweltfragen
Herrn Thomas Goppel, MdL
Postfach 81 01 40
Federal Republic of Germany
Dear Minister Goppel:
Your recent granting of the second partial license for construction of the FRM-II nuclear research reactor at Garching is alarming because the fuel which will represent the bulk of the core -- high density, bomb-grade, highly enriched uranium (HEU) silicide fuel -- has never been tested, qualified, or used as fuel in any reactor. Consequently, this approval violates internationally recognized standards of nuclear safety and raises unacceptable risks of a catastrophic accident and release of radioactivity to the environment, as described below.
We are writing to urge you to reconsider this decision and to issue a "stop work" order immediately, at least until the proposed fuel can be fully tested and evaluated. Such a prudent delay also would provide time for the operator, Technical University - Munich (TU-M), to reconsider modifying the reactor design to use previously tested and qualified, high-density low enriched uranium (LEU) silicide -- an alternate fuel that could address not only fuel-safety concerns but the risks of nuclear terrorism, nuclear proliferation, and inadequate fuel supply raised by the proposed HEU fuel.
Approximately 80% of the FRM-II's core volume will be made of high-density (3.0 g/cc) silicide-based HEU, a fuel that has never before been tested, qualified, or used as fuel in the core of a research reactor. In the 1980s, Argonne National Laboratory tested two miniature plates containing HEU silicide fuel, but the uranium density was only 1.7 g/cc. These irradiations were performed not to qualify an HEU fuel for routine use in a reactor, but only to test the burnup limit of the fuel to assist in the qualification of LEU fuel. Limited irradiations of this type have never been used before as the sole basis for qualifying any fuel or licensing the use of any fuel in a reactor. The only other irradiation tests ever performed on HEU silicide fuel, to the best of our knowledge, were those recently conducted at France's SILOE reactor on behalf of the TU-M, but those were tests of fuel with a density of 1.5 g/cc, only half the density of the fuel which will represent the bulk of the FRM-II core. No tests have ever been conducted on 3.0 g/cc fuel, even though high-flux reactors suitable for conducting such tests are available.
It is highly extraordinary, and may be unprecedented, to license a high-power nuclear research reactor without previously conducting safety tests on the proposed fuel. As noted in a presentation to the 1997 international RERTR conference: "Normal licensing practices in many countries require that tests be performed on the specific fuel that will be used in a reactor in order to provide the data on fuel behavior that is required for licensing." Indeed, until December 1996, the operator of the proposed FRM-II did intend to conduct tests on the proposed high-density (3.0 g/cc) HEU fuel. These plans were abandoned, apparently, when the operator learned that the intended test facility (France's SILOE reactor) was not licensed to withstand the level of heat this fuel would produce in its core. Testing the fuel would have required resort to another reactor, delaying the testing regime, so the tests were abandoned.
It is unacceptable and reckless behavior to abandon safety testing of the fuel simply because such testing is inconvenient and will result in delays in reactor construction. The responsible step is to delay licensing and construction, to provide time for such tests, rather than rush ahead with building and operating a reactor whose untested fuel could result in a catastrophic accident.
The operator did conduct tests on a lower density (1.5 g/cc) HEU silicide fuel, but this will constitute only about 20% of the FRM-II core volume. This test is a start in qualifying the 1.5 g/cc fuel, even though it was conducted for a non-prototypical fission density. However, for several reasons, this test is wholly inadequate to qualify the HEU silicide fuel with 3.0 g/cc that will be used in 80% of the FRM-II core volume. First, the high-density HEU contains twice as many uranium silicide particles per unit volume as the tested fuel. Therefore, there is a greater tendency to swell and less aluminum to hold the fuel together. Second, the neutron flux projected for the FRM-II is more than twice that in SILOE. This means that the fission rate and, thus, the heat generation rate per unit volume projected for the bulk of the FRM-II core will be much greater than those experienced in the test. As noted above, it was this fact that caused the test of the 3.0 g/cc fuel to be abandoned. The 3.0 g/cc HEU silicide fuel needs to be tested under prototypic conditions of fission rate, fission density, and temperature. From a technical standpoint, it is impossible for a licensing authority to certify the safety of a proposed reactor until such tests are conducted.
The potential dangers of operating a reactor whose fuel has not been tested should not be underestimated. One realistic accident scenario raised by the lack of testing is as follows: Under prototypic operating conditions, the untested fuel could swell more than anticipated. If one or two adjacent fuel plates swell sufficiently, they may touch each other, choking off the flow of coolant, which would lead to further excessive heating. If fuel became sufficiently hot, its cladding might be breached, which would bring the fuel "meat" into direct contact with cooling water, releasing fission products into the reactor. Such a medium-level accident would require, at a minimum, the shut-down and decontamination of the reactor.
A more dangerous scenario is also plausible. If under actual operating conditions several fuel plates swelled simultaneously and came into contact with each other, substantially choking the flow of coolant, there could be a core-melt accident, resulting in a catastrophic release of radioactivity to the environment. The probability of such a scenario may be low, but there is no way of knowing with confidence in the absence of tests on the planned fuel.
As noted above, there is an alternative fuel that already has been successfully tested and qualified -- high density, LEU silicide fuel -- which could avoid many of these safety concerns without further testing and which could produce the intended performance in the FRM-II with only a minor re-design. Such fuel already has been used to convert approximately 20 research reactors around the world that originally used, or were designed to use, HEU fuel -- including Germany's FRG-1 and BER-2 reactors (the latter completing its conversion this year) -- and several more conversions are underway. Moreover, according to the most recent expert analysis, conversion of the FRM-II to an LEU design not only would enable equivalent reactor performance, it would actually increase slightly the reactor's fuel-cycle length, a prime objective of nuclear researchers. However, these benefits of conversion to LEU fuel are possible only if the size of the reactor core is increased slightly from the current design. This is further grounds for an immediate halt to any ongoing construction at the FRM-II reactor site.
We wish to underscore that beyond the safety benefits of conversion to previously tested and qualified LEU fuel, conversion to such fuel would bring other substantial benefits. First, avoiding bomb-grade HEU fuel would essentially eliminate the risk of terrorist seizure of the FRM-II's fuel, either in transit or at the reactor site, in order to fabricate a nuclear weapon. Current plans call for the FRM-II to use 1.2 metric tons of HEU over the course of its 30-year life, sufficient for dozens of nuclear weapons. Second, such conversion would bring Germany fully into accord with prevailing international norms that eschew use of bomb-grade uranium in large new research reactors. In the last 20 years, only China and Libya have violated this consensus, and since then, China has joined by designing its next research reactor to use LEU. Conversion of the FRM-II would reinforce the international norm, make it less likely that others will violate it, and thereby help to further reduce and eventually eliminate international commerce in bomb-grade uranium. Third, such conversion would avoid the troubles the operator has encountered in trying to obtain sufficient HEU to fuel the reactor in the face of international efforts to phase-out commerce in such material. Conversion would thus avert the looming prospect of the reactor being completed at considerable expense, only to find insufficient fuel available to enable its intended operation.
For all of these reasons, we call on you to issue immediately a "stop work" order on construction of the FRM-II until, at a minimum, the high-density HEU silicide fuel has been thoroughly tested and evaluated under conditions that realistically simulate those of projected operation of the FRM-II. Moreover, we urge you to direct the operator of the FRM-II, the Technical University - Munich, to re-evaluate the safety and nonproliferation benefits of converting to already tested and qualified high-density LEU silicide fuel.
Thank you for your attention to this urgent matter.
Senior Policy Analyst
Edwin Lyman, PhD
What's New HEU PageHome Page